Medium
Assumptions Testing
SPSS
Free
A student runs a hierarchical regression in SPSS predicting job performance from Step 1: conscientiousness, Step 2: stress, Step 3: conscientiousness × stress (moderation). They center conscientiousness and stress first. SPSS shows:
- Model 3: VIFs are 1.4 (conscientiousness), 1.6 (stress), 1.3 (interaction)
- Condition Index peaks at 12
- ΔR² from Step 2 to Step 3 is significant Which assumption-related conclusion is MOST appropriate?
A
Multicollinearity is a serious problem because interaction terms always cause VIF > 10, so the moderation result is invalid.
B
Multicollinearity does not appear severe; centering helped, and the significant ΔR² supports adding the interaction for moderation.
C
Because ΔR² is significant, normality and homoscedasticity assumptions are automatically satisfied.
D
A Condition Index of 12 proves perfect independence of errors, so no further assumption checks are needed.